Report
on the
Irish Cervical Screening Programme
JULY 2004
Dr. Euphemia McGoogan
Senior Lecturer Pathology, University of Edinburgh
Associate Medical Director, Lothian University Hospitals Trust, Edinburgh
Consultant Cytopathologist, Royal Infirmary of Edinburgh
Report
on the
Irish Cervical Screening Programme
JULY 2004
Dr. Euphemia McGoogan
Senior Lecturer Pathology, University of Edinburgh
Associate Medical Director, Lothian University Hospitals Trust, Edinburgh
Consultant Cytopathologist, Royal Infirmary of Edinburgh
introduction
I was invited by Dr Sheelah Ryan, Chief Executive Officer, on behalf of the Health Board
Executive (HeBE), to undertake a review of the operation of the first phase of the Irish Cervical
Screening Programme (ICSP) in the Mid Western Health Board as an overview. The purpose of
the retrospective evaluation was to identify the improvements that may be needed to move
forward nationally. My help was requested in clarifying the model, particularly in relation to
policy, as the Department of Health in Ireland begins to grapple with the changes in the
organisation of the Irish Healthcare system.
TERMS OF REFERENCE
•
Evaluation of the present organisational elements in Phase 1 ICSP as defined by the
Report
of the Department of Health Cervical Screening Committee published
December 1996
•
Evaluation of the application of quality standards as defined by the Quality Assurance
document of the National Expert Advisory Group on Cervical Screening 1999
•
Review of Phase 1 in the context of preparedness for the organisational issues for a
national rollout
•
Defining the critical success factors for a national rollout of the cervical screening
programme.
METHODOLOGY
Between July and November 2003, I visited Ireland on several occasions in order to meet
stakeholders in the delivery of Phase 1 of the Irish Cervical Screening Programme to review
overall progress and to identify successes and any barriers to roll out;
I met with Dr. Sheelah Ryan;
I met with Dr Marian O’Reilly and Programme Office staff;
I met with Dr Kevin Kelleher, Director of Public Health, Mid-Western Health Board;
I visited the cervical cytopathology laboratories in Galway, St Luke’s Hospital , Dublin and Royal
College of Surgeons of Ireland, Dublin;
I visited the Mid Western Regional Hospital Histopathology Laboratory and the Colposcopy
Clinic in the Regional Maternity Hospital in Limerick;
I had meetings with several hospital managers;
I attended a meeting at the Department of Health and Children (DoHC) on 14th November
2003 at which the Deputy Chief Medical Officer, Principal Officer Secondary Care (Cancer
Strategy), Principal Officer Community Care, Assistant Principal Officer Community Care,
Assistant Principal Secondary Care Colposcopy, among others were present;
I have read many documents including: DoHC Report December 1996, First annual report of
ICSP Phase 1 Mid-Western Health Board (MWHB), Colposcopy Survey 2001, Draft Task Force
Report September 2002, ICSP Cytology Laboratory Survey 2002, Draft Service Agreement with
cytology laboratories September 2002, General Practitioner contract (October 2000), Quality
Assurance Document December 1999, Quality policy statement, ICSP Business Plan December
2003, ICSP Cytology Referral Form, ICSP Women’s Charter and was given copies of the ICSP
Video, and ICSP Leaflets.
I would like to express my appreciation for the warm welcome I have received on each of my
visits. I found all the staff groups with whom I met open and frank and they provided me with
all the information I required. This made my task so much easier.
Thank you to everyone.
contents
Executive Summary
1.
Recommendations
2
2.
Background to cervical cancer
9
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
Incidence and mortality worldwide
Incidence and mortality in Ireland
Demographics
Aetiology
Pathogenesis
Histological types and stages
Survival rates
Terminology
Diagnosis and Treatment of pre-invasive disease (CIN)
9
9
9
9
10
10
11
11
12
3.
Principles of screening
13
3.1
3.2
3.3
General principles
Value for money and equity of access
Organisation of screening programmes
13
14
15
4.
Screening for cervical cancer control
18
4.1
4.2
4.3
4.4
4.5
4.6
General principles
Monitoring and Evaluation
Sensitivity and Specificity
Age to start screening
Population coverage and frequency of screening
Age to stop routine screening
18
19
20
20
21
21
5.
The routine screening test
22
5.1
5.2
5.3
5.4
22
22
22
23
5.5
Conventional cervical smears
Liquid Based Cytology (LBC)
Automation Assisted Screening
HPV testing in primary screening and
triage of low grade abnormalities
Other molecular tests
6.
The roles of the hospital based service
25
6.1
6.2
6.3
Cervical Cytopathology laboratory
Colposcopy
Histopathology
25
25
25
7.
Evaluation of Phase One of the ICSP in MWHB
27
7.1
7.2
The background
Organisational elements, achievements and the
application of quality standards
7.2.1
Programme Director
7.2.2
Central Programme Office / Register
7.2.3
Promotion of programme
7.2.4
Smear takers in Primary Care and other settings
7.2.5
Cytopathology laboratory service
7.2.6
Colposcopy service
7.2.7
Histopathology service
27
29
24
29
29
29
30
30
33
33
contents
8
National rollout
35
8.1
Preparedness for a national rollout and the critical
success factors
8.1.1
Governance structures, management and accountability
8.1.2
Professional direction – national committees
8.1.3
Programme direction
8.1.4
The population Screening Register
8.1.5
Uptake of screening
8.1.6
The screening test
8.1.7
Service quality and delivery
Steps in transition to full programme implementation
Improvements that require to be made before full implementation
8.3.1
Cervical cytopathology service
8.3.2
Colposcopy service
8.3.3
Histopathology service
35
8.2
8.3
Appendix 1 Laboratory capacity calculations
35
36
36
36
37
37
38
39
41
41
42
43
44
executive summary
1
The invasive cervical cancer rate in Ireland is one of the highest in Western Europe. In order to
achieve an 80% reduction in cervical cancer in the Irish population, there is a need for an
organised population health based programme rather than continuing with opportunistic
screening as at present. Phase 1 of the Irish Cervical Screening Programme in the Mid-Western
Health Board worked well and provides a sound basis for national roll out. The main weakness
was governance of the cytology laboratories. A transition period of 18 months or less is
required to prepare for the invitational phase beginning. Work during this period needs to
focus on rationalisation of laboratories, accreditation of laboratories, smear taker training, the
Register and links to colposcopy, all of which are detailed in the text.
The cornerstone of the Programme must be quality assurance with clear governance and
accountability, underpinned by both clinical and operational standards. A national Steering
Group should be set up to oversee and provide professional direction for the Programme.
Strong leadership is needed as is a secure funding policy from the Department of Health and
Children for all aspects of the Programme. The significant resource currently utilised in cervical
screening must be redeployed to provide a world class service for the women in Ireland.
1
recommendations
1. The Programme
1.1
A national Cervical Screening Programme must be rolled out as soon as possible.
1.2
There is currently a significant resource deployed in making the cervical screening test
available to women in Ireland on an opportunistic basis. Unfortunately, this has failed
to achieve a significant reduction in the incidence and mortality for cervical cancer.
Indeed there is a suspicion that the trend is upwards. The current resources must be
reconfigured and change actively managed to deliver a national screening programme
for women so that proper use is made of the all the available resources to ensure a
drop in morbidity and mortality from cervical cancer in Ireland.
1.3
Cervical Screening in Ireland should be managed as a single, national, organised
(call / recall) programme with written standard operating procedures, targets and
standards, clear governance and accountability. Only in this way can the best benefit
be provided for Irish women in terms of a decrease in the incidence and mortality from
cervical cancer.
1.4
The ICSP should be organised around four regional screening centres. A typical
screening centre would serve a defined population through a network of professionals
working in primary care, laboratories and colposcopy.
1.5
Synergies with other screening programmes should be explored. A key element to be
addressed in terms of clarity with regard to governance for the Programme involves
the synergies with Breastcheck in moving towards a public health population based
screening programme model within the health reform process.
1.6
The natural home for a cervical screening programme is within a Population Health
department. Given that the health service in Ireland is undergoing restructuring at
present and the final structure is not yet clear to me, it is difficult for me to
recommend exactly where the Irish Cervical Screening Programme should be located.
However, wherever it is based, it is critical to ensure that the ICSP is appropriately
organised and governed to achieve success.
1.7
A management structure with clear governance arrangements is required to ensure
that the legal obligations to eligible women, health care professionals and staff
participating in the Programme including Phase One are met.
1.8
DoHC directives regarding Programme Policy and Quality Assurance are required so
that clarity and consistency is ensured.
1.9
The establishment of effective governance arrangements with immediate effect is
necessary in order to provide overall leadership and direction in terms of policy
development, standards, accountability and achievement of value for money for Phase
One and the extension of the programme.
1.10
A single central Programme office and Programme Director is required with clear policy
that everyone adheres to, responsibility, ring fenced funding and multi-annual budget
planning. The ICSP Programme Director should have clear and effective lines of
communication to the appropriate Director within the Department of Health and
Children (DoHC).
1.11
A multi-disciplinary national ICSP Steering Committee with representation from
management and the professional academic bodies, is needed to provide general
oversight of the ICSP and effectively influence policy development and issues
recommendations
3
pertaining to the Programme to reduce risk. The ICSP Programme Director should be
advised by the Steering Committee.
1.12
A quality assurance (QA) programme should continue to be given priority in the ICSP.
The QA programme must be updated on a continuing basis and in a formal manner. A
culture of continuous quality improvement with openness and accountability should be
encouraged among all stakeholders. Steps must be taken to improve the validity of the
data used to measure the effectiveness of the cervical screening. The use of
confidential enquiry and audit to identify system failures should be developed.
1.13
There are no structures in place to allow immediate review or revision of the QA
document by a professional group. Given Ireland’s geographic location and population
size and since the Irish cytology and colposcopy services already aspire to the
standards of UK NHS Cervical Screening Programme (NHSCSP) and British Society for
Colposcopy and Cervical Pathology (BSCCP), the ICSP should immediately adopt these
standards as the minimum with review in the light of the experience of the first three
years of invitational screening. Phase 1 ICSP has already aligned in an observer status
at the NHSCSP QA Committee
1.14
Structures must be put in place to allow review or revision of quality standards by a
National ICSP Quality Assurance Committee which should include representation from
the Professional bodies, other key stakeholders such as the Cancer Registry, the Cancer
Society and other cancer screening programmes and, most importantly, women.
1.15
The programme outcome should be monitored using the following critical measures of
success:
• Incidence and mortality from cervical cancer in those women screened and those not
screened. This should be measured accurately and not simply by estimate using (out
of date) census data
• Population coverage –the number of eligible women who have a smear test within
the screening interval and the percentage in comparison to the eligible population
age group
• Cervical cancer rates for women who have had a smear test within the defined
screening interval. Microinvasive cancers should be analysed separately.
1.16
A clearly defined screening policy should be developed and implemented which should
be subject to ongoing review in the light of developing evidence, guidance,
technology and government policy.
1.17
Introduce a 3-year routine screening interval for the national programme from 25
years until 35 years of age; and thereafter a 5-year screening interval for routine
screening of women with 2 consecutive negative smears until the age of 60 years. If
laboratory capacity allows, 3 yearly screening intervals may be extended to the 40 or
even 45 years age group. Thereafter 5 yearly screening is adequate for women who
have no history of abnormality. This screening policy was endorsed by the World
Health Organisation (WHO) and the International Agency for Cancer Research (IARC)
earlier this year.
1.18
Ireland must be ready to “go-live” with the invitational phase within 18 months of the
decision to implement. A transition period of no more than eighteen months is
required before the first invitational is rolled out nationally and the first cohort of
women receives letters. Routine activity should continue in parallel during the
transition phase except for those specified categories of smears that can cease
1
recommendations
immediately. A summary of the key tasks to be completed during the transition phase
are listed in Table 11. Investment is needed to match all of the tasks that must be
completed during the transition period and there should be an agreed business plan.
1.19
Since women can enter the Programme by responding to an ICSP invitation letter or by
direct referral at the discretion of the clinically responsible doctor, consideration should
be given to a rapid national extension with conversion of opportunistic smears to
Programme entry policy. After the transition period, smears taken outside of
Programme policy and all subsequent opportunistic smears should be actively
discouraged so that laboratory capacity can be concentrated on delivering results for
programme smears.
1.20
A single standard screening test needs to be agreed. The Irish Cervical Screening
Programme should move to liquid based cytology (LBC) as the routine screening test
soon as possible. LBC brings an immediate reduction in the workload for smear takers,
laboratories and colposcopy clinics due to a reduction in unsatisfactory smears (which
require an immediate repeat test) and an increase in laboratory capacity since more
slides can be screened each day by laboratory staff. If implementation of LBC is
deemed to require an EC Procurement process, an early decision is particularly critical.
Moving to LBC will require a national training programme for converting laboratory
staff and smear takers to the new technology and this will take many months to
implement. This should be fully implemented before the invitational phase starts in 18
months. This is yet another reason for a rapid decision by the DoHC.
1.21
Processes should be put in place to evaluate new technologies in cervical screening
with respect to the ICSP such as computer assisted microscopy and biomarkers as well
as HPV testing.
1.22
I have considered but decided I cannot recommend the use of HPV testing as the
primary screening test in the ICSP at present. There is as yet insufficient evidence that
HPV DNA detection as a screening test would deliver the same efficacy as cytology in
organised screening programmes. We must await further data from trials in progress
such as the Manchester ARTISTIC trial.
1.23
There is also currently insufficient evidence for the cost effectiveness of reflex HPV
testing of low grade samples. I believe we must await the outcome of the NHSCSP
pilot data and other large studies currently in progress in UK and Europe such as the
MRC TOMBOLA trial.
1.24
Active management of the laboratories is required. This is a big challenge and will
require major re-design. Governance of laboratories was a major issue outstanding in
the Phase One pilot. However, it may be that HIQA would have a role in the external
governance of the transition as well as the Programme as a whole. Agreed standard
SOPs are needed for every aspect of the programme to enable appropriate audit and
measurement against targets and standards.
1.25
Since it is a significant portfolio, laboratory redesign would benefit from the
appointment of a short term (24 months) Project Manager to, among other tasks,
manage the transition of the laboratories to meet the needs of the ICSP. This Project
Manager should report to the Programme Director.
recommendations
5
The Screening Programme Register
1.26
A complete register of all women in Ireland aged 25 – 60 years must be established
using a unique identifier for each woman to minimise the clinical risks of duplicated
records.
1.27
The unique identifier number for women should be the Personal Public Service (PPS)
number as described in the National Health Information Strategy published in July
2004. This should be made mandatory for all laboratory requests, referrals to
colposcopy and further investigation or treatment records in order to ensure all clinical
information is held in a single record for each woman.
1.28
All relevant clinical history should be held on the central cervical screening register
information system. This should be defined, reviewed and audited at regular
intervals.
1.29
A methodology of recording on the Register all previous smear / colposcopy history for
women joining the programme should be established. Everyone should co-operate to
identify women who have been screened in the past. All current database owners
should clean up their data as best they can, especially merging all duplicate records
for women into a single file with her PPS number attached. The issue of consent
requires to be explored with a view to enabling the Programme to access clinical
histories in existing cytology laboratory databases.
1.30
Responding to an invitation or agreeing to have a cervical screening test incorporates
signed consent not only for the test but also full buy-in to all aspects of the
programme. Clear information about the programme must be made available to
women at the time of invitation or taking the test.
1.31
Women can enter the Programme following an ICSP invitation letter or by direct
referral at the discretion of the clinically responsible doctor. Direct referral would only
be appropriate during the transition phase for entry to the Register.
1.32
The information on the Register must be kept completely confidential. However, all
stakeholders (laboratory, Colposcopy Clinic, Primary Care and the Programme Register
Office) should be able to access a woman’s history at specific points in time in order to
ensure the overall cost effectiveness and the affordability of the Programme. This will
avoid the necessity for colposcopists to have a printed copy of the referral smear result
for each patient. Laboratories will be able to see the results of previous smears or
biopsies for the woman that may have been reported elsewhere and give an
appropriate clinical management recommendation.
1.33
IT linkages between laboratories, colposcopy clinics, smear takers and the Register
must be implemented and tested.
1
recommendations
Smear Takers
1.34
Formal registration of smear takers should be implemented.
1.35
There should be no routine annual tests. I recommend stopping immediately
repeating the smear one year after a first test and all routine repeat cytology at the
first colposcopy visit.
1.36
In order to facilitate cessation of opportunistic screening, clinical protocols in
maternity services, gynaecological outpatients and STD clinics should be updated to
remove any reference to taking opportunistic smears once the invitational programme
is rolled out. This is required to free up laboratory capacity to deal with programme
smears and allow equity of access for all women as well as value for money.
1.37
Consideration should be given to carrying out smear taker ICSP QA visits since primary
care staff are pivotal to a successful cervical screening programme.
1.38
Any review of the GMS contract should include consideration of the fee for taking
Programme smears. The success of the Cervical Screening Programme is reliant on
primary care staff actively promoting uptake among eligible women and adherence to
guidelines.
Hospital Based Services
1.39
There must be a decision about the number and location of cervical cytopathology
laboratories, histopathology laboratories and colposcopy clinics so that there is
appropriate capacity for the programme now and in the future. I recommend a model
similar to that for Breastcheck. Services should be based in four centres, two in Dublin,
one in the south and one in the west. The cytopathology laboratory, colposcopy clinic
and histopathology service should work closely together as a regional team to ensure
the catchment population is screened appropriately and quality standards are met.
1.40
There should be direct funding for cervical cytopathology, cervical histopathology and
colposcopy services - either within a hospital setting or, if necessary, as stand alone.
The requirements of the Programme must be clearly defined and costed and funds
earmarked for Programme use.
1.41
All laboratories providing services for the programme should apply for external
laboratory accreditation before January 2006. Application for CPA(UK)Ltd accreditation
means that the laboratory is ready for an external inspection i.e. a quality manager is
appointed, standard operating procedures are written and implemented, all relevant
quality assurance standards are in place and statistical information is available.
The Cytopathology Service
1.42
Each laboratory should serve a defined catchment area so that reporting profiles can
be compared to the incidence of disease in that regional population. In order to plan
for capacity, Primary Care smear takers should be registered on a geographical basis
with a particular laboratory. However, the national screening laboratories must work
together as a team, operating as a single service with standardised protocols to
provide a contingency so that there can be a smooth delegation of work to another
laboratory / laboratories where there is capacity if one laboratory develops
unacceptable turnaround times.
recommendations
7
1.43
During the transition period when liquid based cytology training is being undertaken,
laboratories may utilise the skills of staff in other local cytology laboratories in a hub
and spoke arrangement so as to make best use of equipment and availability of
trained cytology staff.
1.44
National laboratories should comply with all national standards.
1.45
The protocol for internal quality control procedures should be standardised across all
laboratories and implemented urgently to avoid continuation of unnecessary work
within the laboratory. The responsibility for recalling women at the appropriate interval
should be passed to the Register thus freeing up laboratory time for assessment of
slides.
1.46
Only Programme policy smears should be processed at national screening laboratories.
1.47
The national laboratory structure should be based upon the model described
in Table 12.
1.48
I note with some concern that over 25% of staff are aged 50 years and over and are
due to retire in the next 10-15 years. Consideration should be given to improving
recruitment and retention of cervical cytopathology laboratory staff. Innovative ways
to address the discrepancies in the take home wages between cytopathology and staff
in other laboratory disciplines should be explored.
1.49
Consideration should be given to employing Advanced Practitioners to address staffing
issues in cytopathology laboratories.
1.50
Training in third level is required to improve recruitment of medical laboratory
scientists.
1.51
A defined productivity standard per medical laboratory scientist would assist capacity
planning based on projected workload for the catchment area.
1.52
Cervical smear reports must contain a classification using the dyskaryosis terminology,
a recommendation for management and where not negative, a descriptive text.
The Colposcopy Service
1.53
There should be four national colposcopy centres aligned to cervical cytopathology
and histopathology laboratories with geographically spread outreach clinics for easier
access by the population.
1.54
Outreach clinics should be managed by the national colposcopy centre and quality
standards included in those of the centre. All colposcopists should meet the agreed
national standards.
1.55
Colposcopy should be an outpatient service - and women attending should not be
considered as “patients” unless an abnormality if identified.
1.56
It is important that colposcopists meet regularly with the cervical cytopathologists
(and Advanced Practitioners) and histopathologists to discuss the best management of
difficult cases and to review discrepant clinical findings.
1
recommendations
The Histopathology Service
1.57
There should be four national cervical histopathology services aligned to
cytopathology and colposcopy.
1.58
A lead histopathology consultant should be identified in each laboratory with
responsibility for liaising with the ICSP and adherence to quality standards.
1.59
Quality assurance standards should be agreed immediately and ICSP QA visits
implemented. National histopathology laboratories should implement all national
standards.
1.60
All cervical biopsies should be classified using the SNOMED system and forwarded to
the Register.
2
background to cervical cancer
9
2.1
Incidence of and mortality worldwide
Cancer of the cervix is the second most common cancer among women worldwide,
with an estimated 471,000 new cases and 235,000 deaths in the year 2000. Almost
80% of the cases occur in the developing countries, where, in many regions, it is the
most common cancer among women, responsible for about 15% of all new cancers.
Cervical cancer is less common in economically developed countries, where in the year
2000, it comprised about 4% of cancers in women, ranking sixth in importance.
Effective cytological screening programmes have resulted in dramatic falls in incidence
and mortality of cervical cancer rates over the last 20 years with the UK setting an
important example. The very effective call and recall system in the UK, having
operated since 1988, has seen a progressive fall in mortality at the rate of 7% per
annum with over 100,000 lives saved in the intervening period. This is notwithstanding
an increase in pre-invasive lesions over the same period. However, before the
introduction of screening programmes in the 1960’s and 1970’s, the incidence was
much as we see in developing countries today.
2.2
Incidence and mortality in Ireland
In 1995 the age-standardised incidence rate for Ireland was 9.5 per 100,000 compared
to 13.9 per 100,000 in the UK. Despite the differences in incidence, mortality was
almost the same in both countries (4.6 compared to 5 per 100,000 in the UK)
indicating that Irish women were being diagnosed with their cancer at a later stage
when medical intervention was less likely to achieve a cure. Currently 77 women die in
Ireland each year from cervical cancer. Because cervical cancer affects relatively young
women, it is an important cause of lost years of life.
2.3
Demographics
Incidence shows a rapid rise to a peak at a comparatively young age. Average age of
onset worldwide is at about 25 years, there is a rapid rise between 30 and 40 and a
peak at ages 44 to 49 years. In countries with established screening programmes
invasive cancers are identified at an earlier stage and the peak is at 35 - 40 years.
After the peak, the decline is fairly rapid but there is another small peak at 70-75
years. Cervical cancer also has quite marked differences in incidence, according to
demographic variables such as social class, marital status, ethnicity and religion. There
is a consistent association between risk and early age at initiation of sexual activity,
increasing number of sexual partners of females and other indicators of sexual
behaviour. The part played by sexual behaviour and history of male partners has also
been recognised as significant in increasing risk. These findings are strongly suggestive
of a causative role for a sexually transmitted agent. This agent is now identified as the
group of high-risk types of human papillomavirus (HPV). It is quite likely that the
observed demographic variables with risk of cancer of the cervix are very largely the
result of differences in exposure and biological response to HPV.
2.4
Aetiology
Almost all cervical cancer is caused by a persistent infection by specific high risk types
of HPV. HPV is a ubiquitous hardy DNA virus that can infect various surfaces in the
body including the hands (warts), the feet, (plantar warts or verucca) and the larynx as
well as the anogenital tract of both men and women. There are about 100 different
types of HPV and these are divided into groups of low, intermediate and high risk
types. Anogenital HPV infection is usually transmitted during sexual intercourse and is
so common that most sexually active men and women will have experienced an HPV
infection at some point in their lives. The low risk types are usually responsible for
visible warts on the vulva or penis and are not normally associated with the
development of cancer. Most people never know that they have been infected with a
high risk HPV since these infections are usually without symptoms and individuals clear
2
background to cervical cancer
the virus without needing any treatment or experiencing any problems. Only a very
small number of women are not able to clear the virus and the virus may produce
abnormalities in the cells of the cervix. In an even smaller number of cases, the virus
will persist and eventually lead to cervical cancer. However, cancer takes many years to
develop (10 to 15 years) and will only happen if the virus is not cleared during this
period.
Persistent HPV infection leads to two categories of pre-invasive squamous lesions:
productive, self-limited HPV infection (low grade lesions), and lesions with the
potential, if left untreated, to progress to invasive cancer. Persistent HPV infection
produces a large variety of morphologic changes in the cervix readily identifiable by
routine light microscopy and at colposcopy.
While persistent infection with high-risk types of human papillomavirus is necessary, it
is not a sufficient on its own to cause the development of cervical cancer. Co-factors
such as immunosuppression (for example due to smoking or HIV infection) and other
molecular deficiencies are likely to be involved.
2.5
Pathogenesis
Cervical cancer progresses through a number of early stages (cervical intraepithelial
neoplasia – CIN) that are asymptomatic and invisible to the naked eye but can be
identified by cytology and at colposcopy. CIN lesions do not invade below the surface
epithelium and therefore do not threaten the life of the woman. CIN is divided into
three grades (CIN1, CIN2 and CIN3). Low grade disease (CIN1) is now regarded as
almost entirely composed of transient productive HPV infection.
Most CIN1 lesions regress spontaneously and only a small proportion progress to
invasive cancer. It is estimated that about half of CIN3 lesions progress to invasive
cancer if left untreated. Unfortunately we are currently not able to differentiate
between those lesions that will progress to cancer and those that will regress.
In countries with cervical screening programmes, the peak incidence of CIN 3 is
between 25-29 years, a full 10 years earlier than the peak incidence of invasive cancer.
Thus progression to invasive cancer takes 10-15 years from the development of CIN3.
However, it is possible that some lesions may progress more rapidly. While rapidly
progressive cancers are described, these fast-growing lesions probably represent one
extreme of the distribution of progression times rather than being a different type of
disease and are probably not amenable to prevention by routine screening
programmes.
2.6
Histological types and stages of cervical cancer
The majority of cases of cervix cancer are squamous cell carcinomas arising in the
transformation zone around the cervical os. Adenocarcinomas arising in the
endocervical canal epithelium are much less common, comprising around 20% of
cancers in screened populations. Cytological screening, at least as traditionally carried
out with a wooden spatula, has had little effect in reducing the risk of adenocarcinoma
of the cervix since these tumours occur within the cervical canal and are not readily
sampled when scraping the outer cervix and the cervical os.
Cervical cancer is staged according to the extent of invasion of the surrounding tissues
and organs. International criteria have been set by FIGO (International Federation of
Gynaecological Oncology) for measuring the spread of cervical cancers (staging) in a
standard way. See Table 1.
background to cervical cancer
11
Microinvasive squamous cancers are well-defined lesions with only minimal spread
below the basement membrane to the deep tissues. They are generally asymptomatic
and are detected at screening. Since the invading cells are unlikely to have reached
blood or lymph vessels lying in the deep tissues, these cancers have an excellent
prognosis and an extremely low incidence of distant spread or recurrence after
treatment. In well-screened populations, microinvasive cancers form a large proportion
of the total invasive disease and should therefore be considered a success of
screening. This should be taken into account when publishing cervical cancer incidence
rates.
Table 1 FIGO STAGING OF CERVICAL CANCER
2.7
Stage 0
Pre-invasive (CIN3)
Stage 1a
Microinvasive - not visible
Stage 1b
Clinically apparent lesion confined to the cervix
Stage 2
Invasion beyond cervix but not reaching lateral pelvic wall
or lower 1/3 vagina
Stage 3
Invasion lateral pelvic wall / lower 1/3 vagina
Stage 4
Involving bladder+/or rectum or beyond pelvis.
Survival Rates
Stage of disease at diagnosis is generally the most important factor determining the
survival of cancer patients. Diagnosis at a symptomatic stage (Stages 2-4) has a poorer
prognosis (See Table 2). Regular screening generally results in diagnosis at an earlier
stage and a better survival, as does rapid access to appropriate treatment.
Table 2 FIGO (1985) 5 Year Survival
2.8
Stage 0 (CIN3)
100%
Stage 1a
97%
Stage 1b
78%
Stage 2
57%
Stage 3
31%
Stage 4
8%
Terminology for pre-invasive cytology and histology
Historically, the terminology used to classify these pre-invasive cellular changes has
undergone periodic revision to incorporate advances in the scientific, clinical, and
cultural understanding of cervical cancer. Many different classification schemes are in
use worldwide and within Europe and there are, in addition, a myriad of local
modifications to these schemes. Meaningful communication between pathologists and
gynaecologists depends on a clear, shared understanding of the terminology being
used. In the late 1980s, the UK and Ireland adopted the British Society for Clinical
2
background to cervical cancer
Cytology (BSCC) dyskaryosis terminology for cytological smears and the CIN (cervical
intraepithelial neoplasia) terminology for histological lesions (see Table 3). The recently
updated US Bethesda terminology is the classification most used outside UK and
Ireland. The inter-observer and intra-observer variation in the use of all of these
classifications is well recognised. The cytology result broadly indicates the expected
histological grade but may underestimate the severity of the lesion in the cervix.
Table 3 Comparison of terminologies used in cervical cytopathology and histopathology
2.9
DYSKARYOSIS CLASSIFICATION
CORRESPONDING CIN CLASSIFICATION
(Cytology)
(Histology)
Unsatisfactory
--
Negative
Normal
Borderline Changes
--
Mild dyskaryosis
CIN1
Moderate dyskaryosis
CIN2
Severe dyskaryosis
CIN3
?Invasive squamous carcinoma
Squamous carcinoma
Adenocarcinoma
Endocervical / endometrial adenocarcinoma
Diagnosis and Treatment of pre-invasive disease (CIN)
The diagnosis of pre-invasive lesions (CIN) depends on obtaining a biopsy from women
with abnormal cervical cytology, using a colposcope (magnifying binoculars) to define
the abnormal area on the cervix. Once diagnosed, certain management protocols come
into play depending on the grade of CIN and the referral cytology.
Women with low-grade cytological abnormality are often managed conservatively by
follow up since the majority of these lesions regress spontaneously (regression rates of
over 60%). In some healthcare settings conservative management is augmented by an
HPV test at 12 months with a finding of a positive high risk HPV DNA type at 12
months being an indicator for colposcopy referral. Colposcopists are divided whether
or not to treat these low-grade lesions but excision biopsy is usually undertaken if the
lesion persists for several colposcopy visits.
The cumulative evidence indicates conclusively that lesions histologically confirmed as
CIN2 or CIN3 have a progression rate of up to 40%. Therefore, treatment of these
lesions is required and should be instigated soon after these lesions are diagnosed.
Treatment for all grades of CIN is normally performed as an outpatient procedure at a
colposcopy clinic and has minimal morbidity. It involves removing the abnormal surface
epithelium together with some of the underlying tissue. This can be achieved by
several methods including laser excision, loop excision, and diathermy as well as
surgical excision.
Follow up post treatment is required since about 10% of women experience some
recurrence.
3
principles of screening
13
3.1
General principles
In general health professionals and the public have a poor understanding of public
health screening programmes. “Screening” is the use of presumptive methods to
detect unrecognised health risks or diseases in order to permit timely intervention. The
World Health Organisation (WHO) pioneered the development of criteria for screening
programmes. In 1968 Wilson & Jungner defined the aims and principles of screening
that have been the mainstay of research into and the application of screening ever
since.
The Wilson & Jungner principles are listed in Table 4: Ideally all their criteria should be
met before screening for a condition is initiated. These principles bring with them
implications for an ethical approach to healthy individuals participating in the
screening process. In medical practice, the special nature of the relationship between a
patient and her physician has resulted in the need to build up a core of ethical
principles which govern this relationship.
Table 4 Wilson and Jungner. Principles for screening programmes
•
The disease should pose an important health problem for the individual and the
community
•
The natural history should be well understood with a recognisable early stage
•
An appropriate and acceptable screening test should be available and offered at
suitable intervals
•
Treatment at an early stage should be advantageous
•
There should be adequate facilities for the diagnosis and treatment of
abnormalities identified
•
The chance of physical or psychological harm must be less than the chance of
benefit
•
The costs of the screening programme should be balanced against the benefits it
provides.
The benefit of the screening programme should outweigh any physical or psychological
harm (caused by the test, diagnostic procedures or treatment). Even if the screening
test is in itself harmless, a “positive” or “equivocal” result may cause unnecessary
anxiety and the subsequent investigations and treatment can be hazardous. Ensuring
the safety of screening is important because large numbers of individuals will be
screened, creating a potential for greater numbers to be harmed, emotionally or
physically by the process of screening.
An important distinction between screening and other medical care is that the
screening encounter is not normally originated by the individual; rather the provider of
screening (governments, public health units, or individual physicians) initiates the
process. In screening, however, those who are approached to participate are healthy
individuals and most of them never become patients. The provider of screening
believes that as a result of screening, the health of the community will be better. There
is therefore an ethical responsibility on those planning to introduce screening to be in
a position to expect an overall benefit in the community.
Some additional ethical issues are also important. The first is the need to minimise the
harm and anxiety that may affect certain individuals. This requires that the screening
3
principles of screening
methods selected should have the lowest proportion of false-positives possible. The
second is to ensure that a useful remedy is available for all individuals who test
positive. There should be no one for whom a definitive diagnostic test is not available.
If this is not the case, screening will merely generate groups of anxious individuals for
whom there is no benefit.
With these objectives there is a need to ensure that those who have an abnormal
screening test return for a diagnosis and, if found to have a significant lesion, attend
for treatment. A screening programme must ensure that there is sufficient contact with
the individuals being screened to make them aware of the implications of an abnormal
test to ensure appropriate follow up. Provision must be made to ensure that they have
somewhere to return to for further medical advice and if necessary, counselling.
Failsafe follow up procedures must be in place for non attenders.
3.2
Value for money and equity of access
All other options for managing the disease should have been considered (e.g.
improving treatment, providing other services), to ensure that no more cost effective
intervention could be introduced or current interventions increased within the
resources available. The opportunity cost of the screening programme (including
testing, diagnosis and treatment, administration, training and quality assurance)
should be economically balanced in relation to expenditure on medical care as a
whole.
Wilson and Jungner stated that the aim of screening programmes is to sort out those
people who probably have the disease from those who probably do not. A screening
test is not intended to be a diagnostic test. Screening procedures are generally easier
to perform and cheaper than diagnostic procedures but their results require
confirmation through definitive diagnostic tests.
Equity of access to screening services is another important consideration. All those
who stand to gain from screening should have access to the procedure. A screening
service should not be a service that relies on individuals seeking out particular tests or
procedures that they have heard may be of value. Instead, those who organise the
service have an obligation to ensure that those who have not heard of the
test/procedure but who stand to benefit from it are adequately informed and located
to enable them to be screened.
In order to achieve this, adequate staffing and facilities for testing, diagnosis,
treatment and programme management should be available throughout the country
prior to the commencement of a national screening programme.
There should be evidence that the complete screening programme (test, diagnostic
procedures, treatment / intervention, management) is clinically, socially and ethically
acceptable to health professionals and the public.
Public pressure for widening the eligibility criteria, for reducing the screening interval,
and for increasing the sensitivity of the testing process should be anticipated.
Decisions about these parameters should be scientifically justifiable to the public.
principles of screening
15
3.3
Organisation of screening programmes
All national programmes also have certain common values and structures ( see Table 5
and Figure 1)
Table 5 Values common to all nationally managed screening programmes
(Taken from J Muir Gray. Screening in Scotland. NSC Programme Director’s report 20022003)
•
People should be offered the opportunity to enter a screening programme, and
given the necessary information and support to make an informed choice based on
their values
•
Professionals involved in screening programmes need development and support
•
Screening programmes aim to maximise benefit, minimise harm, and make the best
use of the resources available
•
Screening programmes need to work in partnership with related clinical services to
provide a seamless experience for people needing treatment
•
Explicit and valid evaluation is essential both to fulfil a responsibility for
accountability and provide a baseline for quality improvement
•
Programmes are committed to continuous improvement in performance and
standards
•
Research can contribute to quality improvement and should be encouraged.
Nationally managed screening programmes are not directly responsible for treatment
services but because they are identifying healthy people, some of whom will eventually
be referred to treatment services; there is a responsibility to do everything possible to
ensure that treatment services are of high quality.
These stages can be identified in every screening programme and the management of
the programme is primarily responsible for organisation, co-ordination, evaluation and
quality improvement. Quality assurance has four objectives:
• To reduce the risk of error
•
To manage errors competently and sensitively when they do arise
•
To support continuous improvement in performance
•
To set and re-set standards.
3
principles of screening
Figure 1 Structures common to all screening programmes
Identifying the
population and
inviting individuals
Quality
Management
Screening
Diagnosis
Treatment
Features common to all nationally managed screening programmes are listed in Table 6
Screening is a programme, not a test, and screening programmes need high quality
information systems to measure their impact and their quality.
Those responsible for screening programmes have an ethical responsibility to ensure
implementation and maintain quality control of the screening tests and that the
effectiveness of screening programmes is continually monitored. There should be a
plan for managing and monitoring the screening programme and an agreed set of
quality assurance standards.
Table 6 Common features of nationally managed screening programmes
•
Cover a defined population
•
Have a simple set of objectives
•
Develop valid and reliable criteria to measure performance and produce an annual
report
•
Organise quality assurance systems to help professionals and organisations prevent
errors and improve performance
•
Communicate clearly and efficiently with all interested individuals and
organisations
•
Co-ordinate the management of these activities, clarifying the responsibilities of all
individuals and organisations involved
Screening programmes focus on performance measurements as their main indicator of
quality but it is also very important to measure outcome. It is essential to be able to
identify all the individuals affected by a condition for which screening has been
organised including those who did not attend for screening because it may be that the
population that is most at risk is that which is not receiving screening. No matter how
good the quality of the screening programme, its impact on the population will be
significantly reduced if it is not covering a high enough proportion of the population at
greatest risk.
principles of screening
17
For this reason registers play an essential part in programme management and
monitoring. There is a need for population registers to be developed and maintained
and for those offered screening to be given information describing why this is
important. Understandable public concern about confidentiality has led to changes in
data management which have made registers more difficult to organise and sustain.
Each screening programme has to develop training specifically for the professionals
involved in a particular type of screening. There is however, much common ground and
it is important for those professionals involved in more than one programme (e.g.
Primary Care clinicians) to understand the general principles of screening and how the
different programmes relate to one another. In particular it is important to develop the
knowledge and skills of public health professionals to provide generic screening
education resources that could be used for any professional group.
4
screening for cervical cancer control
4.1
General principles
The World Health Organisation (WHO) and the International Agency for Research on
Cancer (IARC) published Guidelines for Screening for Cancer of the Uterine Cervix in
1986. Revised guidelines were written earlier this year and are due for publication in
December. A summary is available in the IARC website (iarc.fr). The European Union
guidelines published in 2003 are broadly similar to the IARC /WHO guidelines. It is now
accepted that an 80% reduction in cervical cancer can be achieved in an organised
call-recall Programme implementing quality standards.
WHO recommended that for cervical screening to be effective, it is especially
important that the programme be organised according to an agreed policy. The
essential elements of such a policy are shown in Table 7
WHO also recommended that cervical screening must be run as an organised system
with objectives, criteria to measure progress towards those objectives, standards and
targets.
The outcome indicators to measure effectiveness of the programme are a fall in
incidence and mortality from cervical cancer. A secondary indicator could be a shift in
the stage of invasive stage and fewer with distant spread round the body. The
detection of high grade CIN may also be used as a surrogate for invasive cancer.
Screening for CIN, the pre-invasive stage of cervical cancer meets the above criteria.
The effectiveness of screening programmes based on cytological smears in reducing
the incidence of and mortality from carcinoma of the cervix has been well established
for some decades as evidenced by experience in Scandinavia in the 1960s and more
recently - and dramatically - in the UK.
As explained above, cervical cancer progresses through a number of early stages in
which some of the cells of the cervix become abnormal although they are not yet
cancerous. Cervical screening is the process of identifying these abnormal cells so that
they can be removed before they become cancer. It is a very important part of a
woman’s routine healthcare and all women between the ages of 25 and 60 with a
cervix should be offered regular screening.
screening for cervical cancer control
19
Table 7 IARC Guidance for Cervical Screening Programmes
CRITERIA FOR A CERVICAL SCREENING PROGRAMME
•
the target population has been identified
•
individual women are identifiable
•
measures are available to guarantee high coverage and attendance, such as a
personal letter of invitation
•
there are adequate field facilities for taking smears and adequate laboratory
facilities to examine smears
•
there is an organised programme for quality control of smear taking and laboratory
interpretation
•
adequate facilities exist for diagnosis and for appropriate treatment of confirmed
neoplastic lesions and for the follow-up of treated women
•
there is a carefully designed and agreed referral system, and an agreed link
between the women, the laboratory and the clinical facility for the diagnosis of an
abnormal screening test, the management of any histological abnormality found
and providing information about normal screening tests
•
Evaluation and monitoring of the whole programme is organised in terms of
incidence and mortality rates at the level of the total target population among
those attending, and among those not attending
•
Quality control of these epidemiological data should be established.
Evidence based information, explaining the consequences of testing, investigation and
treatment, should be made available to potential participants to assist them in making
an informed choice.
The three principal factors influencing how much benefit can be obtained in any
population are
• The proportion of eligible women who are screened. This should be measured
accurately and not simply by estimate using (out of date) census data
4.2
•
The sensitivity of the screening test in detecting high grade pre-invasive disease
(i.e. CIN3 and not CIN1). This is a combination of the quality of smear taking, slide
preparation and laboratory assessment
•
The adequacy of colposcopic investigation and treatment of any disease detected.
Monitoring and evaluation
Cervical screening must be run as a system with objectives, criteria to measure
progress towards those objectives, standards and targets. There should be a plan for
managing and monitoring the screening programme and an agreed set of quality
assurance standards. At the initial roll out of a cervical screening programme, these
can be adopted from similar existing cervical screening programmes. However,
standards and targets need to be reviewed regularly in the light of experience and
reset for the individual programme.
The effectiveness of the screening programme is the degree it achieves its objectives;
the quality of the programme is the degree to which it conforms to pre-set standards
of good screening.
4
screening for cervical cancer control
At a national level changes in cancer incidence and mortality among those attending
and among those not attending are valid measures of effectiveness. For the population
served by health authorities or individual laboratories, however, incidence and
mortality data are less valid. Other measures of the cervical screening process, for
example, the percentage of the population who have had a screening test, or the false
positive rate have to be used. A common dataset for the programme evaluation should
be agreed and an annual report should be produced.
Any impact on incidence and mortality will take several years to appear. Incidence may
take up to two screening rounds (6-10 years) to show its full effect. Mortality may take
much longer since women dying from cervical cancer in the first 6-10 years will
generally already have developed their cancer and would not have benefited from the
screening programme.
Steps must be taken to improve the validity of the data used to measure the
effectiveness of the cervical screening. The use of confidential enquiry and audit to
identify system failures should be developed.
Cancer Registry data needs to be linked with other measures of programme
effectiveness.
To measure and improve the quality of the programme, it is necessary to have not only
explicit standards but also an information system which allows the necessary data to
be collected and a quality assurance system which allows action to be taken should
any programme fail to meet those standards.
Quality assurance must be run as a national system so that the same standards,
definitions and data are being used across the whole country. A national Quality
Assurance Team with relevant professionals should be established to ensure
appropriate standards are set. There should be a culture of transparency, openness
and accountability.
4.3
Definitions of sensitivity and specificity
Sensitivity is the ability of the test to detect women who harbour a significant
abnormality in the cervix while specificity is the ability of the test to correctly identify
normal women.
Both are important in screening programmes. The higher the quality of the programme
the higher its sensitivity and specificity. Attempts to increase sensitivity (reduce false
negatives) invariably result in decreasing specificity (increasing false positives
including unsatisfactory or borderline smear rates). Thus attempts to reduce the
number of false negatives may lead to more normal women being recalled for repeat
tests or referred to colposcopy.
4.4
Age to start screening
The incidence of carcinoma of the cervix is very low under the age of 24 years. In
Europe, age to start screening varies widely, with Finland and Holland inviting women
to their organised programme from the age of 30. The NHS Cervical Screening
Programme in England recently recommended that women under the age of 25 should
not be screened. For developing countries, screening may commence at age 35 years in
order to maximise best use of resources since invasive cervical cancer is infrequent
below the age of 35 years. WHO / IARC do not recommend screening below the age of
25 since this is not cost effective.
screening for cervical cancer control
21
4.5
Population coverage and frequency of screening
The design of a screening programme defines two key parameters for achieving these
objectives, coverage of the target population and the screening interval. Compliance
with these parameters is crucial in maintaining the effectiveness of the programme
and in measuring cost-effectiveness in order that appropriate resources may be
directed to increasing population coverage. The potential percentage reduction in
cumulative incidence can only be obtained if a high proportion of the population (over
80%) comply with screening. Significant deviation from the recommended screening
interval or target population may reduce programme effectiveness either by using
excessive resources as in the case of annual re-screening for CIN or by allowing the
disease to ‘escape’ the period at which early intervention can lead to treatment and/or
cure.
The optimal screening interval is one that provides the most favourable ratio between
degree of disease control and cost of screening. Determining the frequency of
screening is aided by an understanding of the natural history of cervical cancer,
including the duration of the asymptomatic (CIN) phase (10-15 years). The
International Agency for Cancer Research and WHO in 2004 recommended a screening
interval of 3 years in younger women since there is almost as much benefit from
screening three yearly as for annual screening. Since the protection from a cervical
smear is higher in older women, the routine screening interval can be increased to 5
yearly after the age of 35 years if the previous two screens are negative.
More rapidly progressive lesions do occur but it appears unlikely that the majority of
such lesions would be detected by more frequent screening.
4.6
Age to stop screening
Invitation for routine cervical screening in women who have been actively screened
and always been negative usually stops at age 60-65years. The European guidelines
recommend 64 as the upper age limit for routine invitation.
There have been suggestions that women who have never had an abnormality on
cytology and have been active participants in screening could stop screening at 55
years or even 50 years but there is no good evidence to support this strategy at
present.
Since the incidence of cervix cancer in all countries shows relatively high rates of
invasive cancer in older women, there is a consensus that women over the age of 60
who have never been screened or have not been screened for many years should be
encouraged to have at least 2 smears, and only if both are negative should they stop
screening.
5
the routine screening test
5
The Wilson & Jungner principles of screening require that there should be a simple,
safe, precise and validated screening test. The test should be acceptable to the
population. The distribution of test values in the target population should be known
and a suitable cut-off level defined and agreed for “positive” results. There should be
an agreed policy on the further diagnostic investigation of individuals with a
“positive” or persistent equivocal and unsatisfactory test results.
5.1
The Conventional cervical smear
The conventional cervical smear (Pap test) has been used as the routine screening test
for many decades and has been responsible for the dramatic reductions in invasive
cancer in well screened populations. It consists of scraping the cervix with a suitable
sampler and spreading the material obtained onto a glass slide then fixing it with a
preservative fluid or spray. It is well recognised that the sensitivity of this method is
only around 55% mainly due to sampling and preparation issues. It is the regular
screening during the long period that CIN3 is present before cancer develops accounts
for the success of cervical screening programmes. The reasons for this low sensitivity
are well understood and include limited transfer of cellular material onto the glass
slide so that the smear may not be representative of the material removed from the
cervix.
5.2
Liquid Based Cytology (LBC)
More recently cervical screening programmes across the world are introducing a new
technology for collecting cellular material from the cervix and preparing slides. This
new technology is called Liquid Based Cytology (LBC) since the cellular material
removed from the cervix is placed into a transport fluid and sent to the laboratory as a
cell suspension. The two main commercial LBC systems currently available on the
market are ThinPrep (Cytyc, Boxborough, MA, US) and SurePath (TriPath Imaging Inc.).
These utilise very different technologies and have differing resource requirements. The
collection methods also differ between the systems.
While the laboratory consumables are more expensive than those for the conventional
smear, the evidence is that LBC markedly reduces the unsatisfactory rate while
increasing or at least maintaining the identification of high-grade lesions. Thus there
are fewer repeat tests and more appropriate referrals for colposcopy. In addition, the
laboratory throughput may increased by about 40% allowing greater productivity.
The recent pilots in England showed robust evidence for cost effectiveness in terms of
life years saved. The National Institute for Clinical Excellence has recommended that
the NHS Cervical Screening Programme in England and Wales convert to LBC for its
routine screening test. Scotland took a similar decision in 2002 and is now fully
converted to ThinPrep as its routine screening test. Scotland chose to implement only
one LBC system to facilitate training and external quality assurance, allow more cost
effective national procurement and to avoid confusion among smear takers.
5.3
Automation Assisted Screening
Automation assisted screening (using computers to scan the slides) is aimed at
increasing sensitivity (and specificity) by identifying abnormal samples especially those
with only a few or very small abnormal cells. It has proved extremely difficult to train
computers to identify abnormal cells on conventional smears.
Two commercial systems were extensively studied in the 1990s: PAPNET (Neuromedical
Systems Inc. (NSI), Suffern, New York, USA) which has gone into receivership and the
AutoPap System (Neopath Inc., Redmond, Washington, USA). Newer devices are
emerging targeting liquid based cytology smears. The ThinPrep Imager (Cytyc,
Boxborough, MA, US) and FocalPoint (Formerly AutoPap) (TriPath Imaging Inc.) have
the routine screening test
23
been approved by the FDA in the US for routine screening. Each system is designed to
work on its own method of LBC preparation only. The ThinPrep Imager scans the slide
and selects the 20 most suspicious fields. These fields are presented down the
microscope to the cytotechnologist reviewing the slide. The FocalPoint system after
scanning the slide presents the laboratory with a risk assessment of the slide
containing abnormal cells. The FDA has approved the system at a 20% sort rate. This is
the 20% lowest risk slides can be archived without human review. The remaining slides
are subject to full manual review. A new version of the TriPath system with the ability
to relocate suspicious fields down the microscope (FocalPoint GS) has not yet received
FDA approval.
The few randomised prospective studies and also the other performance studies have
shown that automation assisted screening may be feasible as a part of routine primary
screening. It appears to perform at least equally well compared to manual screening in
organised well working screening programme. It is estimated that these devices would
increase throughput of primary screening by up to 100%.
5.4
HPV testing as a primary screening test and for triage of low grade
abnormalities.
HPV testing can be performed on the residual material in the LBC vial. There have been
suggestions of using testing for the presence of HPV DNA as a more sensitive primary
screening test for the presence of high grade CIN since the negative predictive value is
very high. Unfortunately, the high frequency of positive HPV tests among younger
women results in poor specificity (many normal women will receive a positive result).
Thus HPV testing as a primary screening method is unsuitable for use under the age of
30 years of age.
The specificity of HPV testing is considered as too low for use as a primary screening
test in populated-based screening since HPV infections most often clear spontaneously.
HPV testing is likely to result in a considerable number of women who do not harbour
high-grade disease being referred for colposcopy and biopsy. Counselling and
management of normal women with persistent HPV positive tests but who show no
evidence of CIN on cytology or at colposcopy is a major problem especially since only a
tiny percentage of such women will ever progress to cervical cancer. Currently there is
no treatment available for persistent HPV infection and the result of this screening test
could simply be unnecessary anxiety.
Before HPV can be considered for primary screening of women over 30 years of age
several issues would require to be addressed including cost, widespread dissemination
of information about HPV, and the appropriate screening algorithm for women younger
than 30 years. Costs of HPV testing are currently higher than for cytology but could
potentially be balanced by increasing the screening interval for HPV negative women.
The Manchester ARTISTIC trial is designed to answer some of these questions.
However, using HPV DNA testing to triage those LBC samples that show a low grade
cytological abnormality may help identify those women who are normal and can be
returned to routine screening. This makes use of the negative predictive value for a
HPV test which is extremely high. A study in the UK on women over 30 years
suggested that a combination of HPV testing and cytology could support longer
screening intervals for those negative on both. The MRC TOMBOLA trial is designed to
answer these questions.
HPV testing is also being studied to see if women who test negative following
treatment for CIN can be returned to routine screening earlier than on current
protocols.
5
the routine screening test
5.5
Other molecular tests using markers of disease progression
Much recent work has gone into identifying test systems potentially more specific than
HPV that might indicate early evidence of progression to high-grade CIN or cancer.
Such tests may allow selecting women at increased risk for cervical cancer with a high
predictive value. Certain bio-molecular pathways, subsequent to HPV infection, that
lead eventually to the development of CIN3 and cancer, are investigated and some
potentially useful molecules have been identified (for example p16 ink4a and mcm5).
The potential advantages of molecular markers in future clinical practice include more
accurate and reproducible classification of histological cervical lesions; clearer
distinction between cervical and endometrial lesions; better prediction of prognosis
and the selection of best treatment procedures. Ultimately these molecular markers
may be used in triage of women with minor cytological abnormalities and may even be
used for primary screening for cervical progressive cancer precursors.
However, the detection and utilisation of such molecular markers is still in the research
stage.
6
the role of hospital services in screening
25
6.1
Cytopathology service
The role of the cytopathology laboratory is well described in the 1996 DoHC document.
Cervical smear test reports include a coded result, a recommendation for when the
next smear should be taken or clinical management and where not normal, a
descriptive text. In order to recommend the appropriate next management of women,
access to accurate information about previous screening tests and clinical history are
necessary.
A new grade of non medical laboratory staff - the Advanced Practitioner in Cervical
Cytopathology has recently been introduced. This grade is recognised by the Royal
College of Pathologists and the Institute of Biomedical Scientists in the UK. Medical
Laboratory Scientists must undergo rigorous training and sit an examination before
becoming qualified to apply for an Advanced Practitioner post. Staff employed on this
grade must work under the general supervision of a consultant pathologist but they
can report abnormal smears and give clinical advice about management of women
with abnormal results. Advanced Practitioners have been successfully introduced in the
UK and have offset the major shortage of consultant pathologists and proved to be
value for money.
CPA(UK)Ltd laboratory accreditation is mandatory in the UK for all cytopathology
laboratories. Regular (at least annual) proficiency testing (a form of external quality
assurance) organised at a regional (national in Wales and Scotland) is also mandatory
for all medical and technical staff reporting cervical smears. Rapid review of all
negative and inadequate slides is the standard internal quality control method and the
sensitivity of all screeners must be calculated according to a defined protocol and
documented. For pathologists, the positive predictive value of smears reported as high
grade (i.e. subsequently confirmed as high grade on histology) is the quality assurance
method of choice. The documented outcomes of internal quality control and external
quality assurance as well as the reporting profiles of the laboratory are all monitored
at regular intervals during the Quality Assurance Team visits.
Liquid based cytology has replaced the conventional cervical smear as the routine
screening test. Implementation is complete in Scotland and is in the process of being
rolled out in England and Wales.
6.2
The Colposcopy Service
Colposcopy is the diagnostic investigation of women with abnormal cervical smears or
persistent unsatisfactory or borderline smear results. The role of the colposcopy clinic
is well described in the 1996 DoHC document. The British Society for Colposcopy and
Cervical Pathology (BSCCP) has subsequently introduced quality standards for
colposcopists including a minimum workload for each colposcopist in terms of number
of new cases each year and an upper limit to the percentage of biopsies reported as
within normal limits. Nurse specialists have also been introduced in some clinics as a
cost effective means of supporting gynaecologists in the routine colposcopy clinic
workload.
It is important that colposcopists meet regularly with the cervical cytopathologists (or
Advanced Practitioners) and Histopathologists to discuss the best management of
difficult cases and to review discrepant clinical findings.
6.3
The Histopathology Service
The role of the histopathology laboratory is also well described in the 1996 DoHC
document. Successful cervical screening requires good diagnostic pathology for the
cervical biopsies taken at colposcopy. It is essential to have a definitive diagnosis
6
the role of hospital services in screening
before any decision can be made regarding the optimal management for a woman
with an abnormal screening test.
The pathologist is responsible for prompt, accurate reporting of cervical tissues
obtained through biopsy and surgical procedures. Each pathology report must
communicate information sufficient to ensure that each patient receives appropriate
clinical management. Standard (CIN) terminology should be used.
The results for all cervical biopsies including the cervical histology result from
hysterectomy samples should be sent to the screening Register to allow appropriate
follow up, cessation of recall where appropriate and statistical analyses.
Furthermore, due to the critical importance of quality assurance to successful cervical
cancer prevention, the pathologist is also responsible for detecting and analysing the
errors that routinely occur in screening programmes by comparing histological findings
to those from previous screening tests. This process, termed cytological/histological
correlation, is required to confirm that women in target demographic groups are
receiving appropriate clinical management.
7
evaluation of phase one of the icsp in the mid-western
health board area
7.1
The background
The national Irish cervical screening programme commenced with a phase one
development of the programme following the report of the Department of Health
Cervical Screening Committee (Dec. 1996) that concluded that cervical screening is a
worthwhile preventive health measure when delivered as an organised screening
programme.
The first phase of the Irish Cervical Screening Programme was established in the MidWestern Health Board in October 2000 with an aim to develop and implement a
population based, organised, call / recall cervical screening programme in a defined
area and to test all of the operational issues before implementing the national
programme (see Table 8). The eligible population was defined as women aged 25 to 60
years in the Board area and the screening interval was set at 5 years.
An Expert Advisory Group on Cervical Screening was established in 1997 and its terms
of reference are listed in Table 9. A Steering Group established in January 1999 to
oversee the implementation of the programme in the Mid-Western Health Board with
representation from the Board and the Department of Health and Children community
health and secondary care divisions. This Steering Group has stood down now that its
remit has been completed.
Table 8 Operational Issues that required to be addressed (from DoHC Report 1996)
1. Development of a Population Register
2. The ICSP computerised register Information System
3. Administration of call recall and GP payments
4. Health Promotion
5. ICSP Documents
6. Links with the Cytology Laboratories
7. Links with smear takers
8. Links with the Colposcopy Clinic
9. Links with the Histology Laboratory
10. Quality Assurance
A national QA document drawn up by the Expert Advisory Group was published in
1999. The document is divided into four components (Table 10)
Following the establishment of the Phase 1 ICSP, the Expert Advisory Group identified
the urgent need to address the permanent organisational and accountability structure
requirements for the national programme. The Department of Health and Children
subsequently requested the CEOs of the Health Boards in August 2001 to examine the
feasibility and implications of a roll out of the programme, such examination to include
consideration of governance arrangements for the programme.
The stakeholders in the Programme included the primary care doctors and nurses;
27
7
evaluation of phase one of the icsp in the mid-western
health board area
cervical cytopathology services; the colposcopy investigation and treatment service
and the histopathology diagnostic service.
The screening test used in Phase 1 ICSP was the conventional cervical smear test and
cervical smear reporting was provided mainly at University College Hospital Galway
and St Luke’s Hospital, Dublin cytopathology laboratories. The Royal College of
Surgeons of Ireland laboratory assisted with a small pilot of liquid based cytology
(ThinPrep) samples in 2003. Cervical biopsy reporting (histopathology) was undertaken
by the Mid Western Regional Hospital Histopathology Laboratory in Limerick.
Colposcopy services were provided by the Regional Maternity Hospital Colposcopy
Clinic in Limerick.
Table 9 Terms of Reference of the Expert Advisory Group on Cervical Screening
TERMS OF REFERENCE
•
To advise the Minister on:
- The piloting of the programme
- The resolution of any difficulties which may emerge
- Developments and best practice including quality assurance and new Technologies
in cervical screening, treatment etc.
- The development of protocols where appropriate
- The setting up and implementation of the national cervical screening programme
- Progress in the implementation of the recommendations of the Report of the
Department of Health Cervical Screening Committee.
•
To support the persons responsible at regional level for the cervical screening
programme
•
To liaise as appropriate with the National Forum on Cancer and with the health
boards in relation to the development of the national cervical screening
programme.
Table 10 The components of the national QA document 1999
•
Quality Assurance Targets for the General Practitioner & Smear Takers
•
Quality Assurance Targets for the Laboratory
•
Quality Assurance Targets for the Colposcopy Service
•
Quality Assurance Targets at the Register Office
evaluation of phase one of the icsp in the mid-western
health board area
7.2
Organisational Elements, Achievements and the Application Of Quality Standards
Overall the Phase one pilot has proved very successful. I am impressed with the
planning, implementation and achievements to date.
7.2.1
Programme Director
The success of Phase 1 has been due in no small measure to the appointment of a
skilled public health consultant as Programme Director. Dr Marian O’Reilly has a good
understanding of screening and she has built up an efficient team in the Central
Programme Office as well as building up good relationships with the external
stakeholders.
7.2.2
Central Programme Office / Register Office
The target population Register has been established, as many duplicate records as
possible eliminated and procedures established to maintain the Register. To date about
87,000 women have their demographic details incorporated into the Register
representing approximately 90% of eligible women (inevitably some women under 25
years and women who are now over 60 years are included on the Register and there
are some duplicate records for which there is insufficient or erroneous data and thus
cannot be merged). Thus Phase One has been extremely successful in establishing the
Register. A cervical screening database containing test results and when the next
smear is due or clinical referral recommendations (clinical management) has been
implemented. A computerised information system for call/recall and failsafe follow up
has been developed and tested. Once registered with the Programme, women are
automatically monitored by the call/recall and failsafe follow up systems to ensure that
they return for a repeat test at the appropriate interval.
Information Technology has been developed to meet all Programme standards and
appropriate support has been established. Information Technology data quality has
been implemented. Data quality has been markedly enhanced through the inclusion of
the Personal Public Service number for women as their unique identifier. This was
viewed as a significant risk reduction measure.
Purchase, supply and distribution of smear test consumables and disposable speculae
to primary care has been organised.
QA of all aspects of the programme has been an integral part of Phase 1 ICSP from the
start and the Cervical Screening Register Office / Programme Office has met all the
standards set in the 1999 QA document but these now require revision. A QA Manager
was appointed and has led the implementation of ISO standards for administration
throughout the Office. The QA Manager attends the English NHSCSP QA meetings in
observer status. The central office achieved ISO 9001-2000 standard of the National
Standards Authority of Ireland in March 2004.
7.2.3
Promotion of the Programme
By May 2004, uptake in the eligible cohort in Phase 1 had reached 71.5% within the
previous five years (with only 3.5 years of Phase 1 completed). Entry to the Register
has been via a number of routes: women responding to an invitation letter from the
central administrative office in Limerick, attending their General Practitioner, attending
a colposcopy clinic or referred for entry to the Programme at the discretion of another
doctor. There has been significant promotion of the programme at smear taker level.
Almost 30% of women responded to their invitation letter in a timely manner – a
commendably good result at the start up of the pilot programme. Revisions to the
wording of invitational letters should be tested and several examples piloted to find
the most effective format for first invitation and recall letters.
Literature has been developed and distributed nationally. Information and awareness
campaigns targeting women are ongoing.
29
7
evaluation of phase one of the icsp in the mid-western
health board area
7.2.4
Smear takers in Primary Care and in other settings
A contract for General Practitioners and Family Planning / Well Women doctors as
independent smear takers has been implemented with a fee per item payment system.
The current fee is ¤46.33 per programme smear. Nearly ¤2 million was paid to doctors
registered with the programme in the first three years. In 2002, 27% of smears were
taken by primary care nurses and this had risen to 35% in 2003 and this trend is likely
to continue in the future. The trend in the UK has been to a nurse led service with over
70% of smears being taken by nurses. Given that contracts are with General
Practitioners, the governance needs to be reviewed in the context of future contracts
since smear taking until now has been perceived as a medical function.
A registration procedure for doctors and nurses as smear takers is in place. A smear
taker training programme, targeting inexperienced smear takers and certified by RCSI
in 2003, has been implemented and a resource manual published. The Nursing and
Midwifery Development Units in some health boards are taking on the running of this
training in their area. ICSP is working with ICGP to develop a model training
programme for smear takers with a view to incorporating a distance learning module.
There have been no Smear Taker QA visits to date and no lead clinician has been
appointed. There has been no progress to establish clinical audit of the programme at
the primary care level nor has audit of the screening history of women who develop
cervical cancer been implemented.
There are many other smear takers in gynaecological outpatients and maternity
services and in Sexually Transmitted Diseases clinics. Taking a cervical smear appears
to have been included in routine clinical protocols preceding the Phase 1 Programme
and should now be discontinued. While these services are a useful means of enrolling
women onto the Register, further opportunistic smears must be avoided in the future.
Once the invitational programme is commenced, opportunistic smear taking must be
removed from clinical protocols in maternity services, Gynaecological outpatients and
STD clinics otherwise unnecessary repeat smears from these sources will swamp the
laboratory capacity.
7.2.5
Cytopathology laboratory services
The Cytopathology laboratory service has a complexity of funding streams (consultant
appointments, cancer services, the ICSP, etc). The Department of Health and Children
also fund the laboratory service providers through the acute hospital services. While
these sources are hugely important, the professional development of the laboratory
needs to be co-ordinated in a way that supports the needs of the Programme.
Laboratories are managed locally within the hospitals and a local laboratory
management committee drives quality assurance standards and coordination with the
ICSP.
Almost 57,000 smear tests from 30,000 unique women were processed in the first three
years. UCHG has processed 37,220 and St Luke’s Hospital 14,148 smear tests in the first
three years. 50% of tests were annual repeat tests following first smears in accordance
with the original policy.
The RCSI cytopathology laboratory participated in an operational trial of 5,000 liquid
based cytology preparations (ThinPrep) in 2002-03 and reported on costs and logistics.
A contract was developed with an external fully accredited laboratory to take on
cytology screening to deal with backlogs that developed.
evaluation of phase one of the icsp in the mid-western
health board area
The first baseline survey of the cytology laboratories in Ireland was carried out in 2002.
This provides a clear overview of the national situation. There were 14 laboratories
reporting cervical smears in Ireland in 2001, half of the laboratories were in the Dublin
area, three in the south and two in the west. Only 5 laboratories had an annual
workload over 15,000 smears per annum minimum standard and between them they
reported over 70% of the national workload. (3 in ERHA and one each in SHB and
WHB). Only three laboratories had a workload over 25,000 smears per annum. 15% of
the workload nationally is reported in overtime.
Staffing complements were difficult to assess since some pathologists worked in more
than one hospital and cytology staff often had duties in adjoining histopathology
laboratories. Better data collection and analyses is required.
Of the data available for pathologists, 10 out of 19 reported less than the 500
minimum smear reports per annum (mainly as a function of the small workload of
some laboratories). Likewise for medical scientific staff the data available shows that
few screen the minimum number of 5,000 conventional smears per annum and the
range was estimated at between 122 and 4989 smears per annum. Training and
continuing education standards appear to be well met in general. The age profile of
medical scientific staff is a concern since over 25% of staff are due to retire in the next
10-15 years.
There is no standard practice at present for clinical management recommendations for
the interval till the next routine smear. Only 2 laboratories followed the 1996 guidance
for a 5 yearly interval and some laboratories chose not to give any recommendation at
all. Most laboratories request an annual repeat smear after the first smear received
from that woman in their own laboratory since there is no way of accurately identifying
whether a previous smear had been taken elsewhere.
Qualitative data about the internal quality control procedures and outcomes is less
well documented. Internal quality control processes are reported to be in place but
these are varied and most are inadequately documented. Individual screeners'
sensitivities were only available in 4 laboratories. The most common method of
internal quality control is rapid review of all negative and inadequate smears. This
however, is complicated by the myriad of double screening protocols in place. Most
laboratories try to undertake biopsy / smear correlation but this is nearly impossible if
the woman did not attend the colposcopy clinic attached to the same hospital. The
positive predictive values (PPV) for smears reported as moderate and severe
dyskaryosis were therefore mainly estimates and individual pathologist’s PPV not
calculated. Calculation of PPV is required for laboratory accreditation and this issue
needs to be resolved through the implementation of defined regional services.
Most laboratories participate in the Irish Association for Clinical Cytology external
quality assurance scheme. The results for this are confidential and assessed at
laboratory rather than individual level. A technical (staining quality) external quality
assurance scheme is also in operation and most laboratories participate.
The quality of laboratory accommodation varies between hospitals.
Some QA standards for cytopathology laboratories were defined in the 1999 QA
document. Subsequently NHS Cervical Screening Programme performance indicators
for laboratories have been adopted. A QA lead for cervical cytopathology has been
appointed on a consultancy basis to carry out QA visits to service providers and
monitor performance indicators and to identify to management gaps in defined QA
standards.
31
7
evaluation of phase one of the icsp in the mid-western
health board area
QA visits have commenced but have proved more difficult to implement than those in
colposcopy. Prior to the visit by the ICSP QA Team, a questionnaire is sent to the
laboratory to complete and a protocol for the visit agreed. Management are included
in these visits so that any resource needs identified for that service can be brought to
the attention and agreed with the hospital managers. This could be a very positive
influence. Following the visit, a draft report is sent to the laboratory for comment
before a final agreed report is issued. To date no final report has been signed off. In
some settings commercial interests for private cytology may be a disincentive to
sharing qualitative QA data.
Strong professional leadership will be required to create a culture of openness and
transparency with a focus on laboratory quality assurance, particularly external quality
assurance. I gained the impression that morale in the cervical Cytopathology
laboratories was low but I saw evidence of real commitment to provide as good a
service to women as possible and a genuine willingness to change in order to achieve
this. Active leadership is required nationally and within laboratories to make this
happen. There appears to have been a lack of confidence among laboratory staff and a
degree of apprehension about the ICSP QA Team visits. However, this is a critical step
to ensure overall quality and safety of the Programme and is one of the most
fundamental aspects still to be addressed.
Historically, laboratories have usually been poorly equipped, staffed and resourced and
have had to fight to retain what little resources they had. Cytology staff do not have
the opportunity to earn additional money through on-call payments that are available
to staff in other laboratory disciplines and therefore staff recruitment and retention
has suffered. Unfortunately this might also encourage backlogs with a need to do
overtime to supplement their basic salary.
Many laboratory practices are out of date resulting in duplication of effort and
exacerbating build up of backlogs. Practices to deal with historical situations are often
continued indefinitely and add to the day to day work of the laboratory. In particular,
much time and effort goes into local failsafe follow up procedures.
Staff are dedicated to delivering a good service for women and many of the redundant
practices have resulted from attempts to improve the sensitivity of smear results being
issued. For example, in some laboratories a large proportion of slides may be fully rescreened two, three or four times before the result is issued. This includes negative
smears. The proportion of smears double screened ranged between 10 and 100%.
Some laboratories fast track (give priority) to smears received from certain types of
clinic or women (for example repeat smears at colposcopy from women referred
following an abnormal smear). This can inappropriately further delay the reporting of
routine smears. Waiting times for smear results (as at 01/12/03) ranged from 1 to 13.5
weeks with only 5 laboratories meeting the national standard of less than 4 weeks.
Most laboratories use the conventional cervical smear but some laboratories in Dublin
have implemented LBC using the ThinPrep system.
The ICSP has encouraged the cytology laboratories to seek formal external laboratory
accreditation with an agency such as CPA(UK)Ltd. Application for CPA(UK)Ltd
accreditation means that the laboratory is ready for an external inspection i.e. a
quality manager is employed, standard operating procedures are written and
implemented, all relevant quality assurance standards are in place and statistical
information is available. Unfortunately I gained the impression that laboratory staff
believe that such laboratory accreditation is unachievable in the short term. There has
been a lack of available resource to complete the documentation required, achieve the
evaluation of phase one of the icsp in the mid-western
health board area
quality standards and failsafe procedures to meet CPA standards before an application
can be sent for accreditation. Only a few laboratories have applied for accreditation
and no laboratory has been successful in achieving accreditation to date.
The workforce is generally very highly qualified. However, there are recruitment and
retention issues and some laboratories are carrying unfilled vacancies. No Advanced
Practitioners are in post.
There has been no accountability of laboratories or the relevant hospital managers to
the ICSP. There have been delays in filling funded vacancies and budget under-spends
are not returned to the ICSP. Hospitals and laboratories are reluctant to have any gaps
in service delivery or standards made public.
7.2.6
Colposcopy Service
One clinic in Limerick acted as service provider for the Mid-Western Health Board area.
It has 1,000 attendances including 380 new patients per year. The Department of
Health and Children fund the colposcopy service provider through the acute hospital
services. A local committee drives coordination with the ICSP.
A colposcopy computer system specification for all clinics was delivered in 2002.
The first baseline survey of the colposcopy clinics in Ireland was carried out in 2001 to
determine the national situation.
QA standards have been defined for the colposcopy clinics. There has been agreement
among colposcopists to align with the British Society for Colposcopy and Clinical
Pathology (BSCCP) standards.
A QA lead colposcopist was engaged on a consultancy basis to carry out QA visits to
service providers and to identify to management, gaps in defined QA standards. Eleven
of the nineteen colposcopy clinics have received a QA visit to date. These visits appear
to have been well received. There is little or no duplication of service and staffing and
these clinics appear to be value for money.
Performance indicators are monitored to meet QA standards. Equipment, clinical
standards and relationships with hospital management were viewed as good but
opportunities for improvement were identified in staffing, facilities and external
communications. Subsequent to the QA visits to the Dublin clinics, the ERHA planners
have rationalised colposcopy services in Dublin. I understand that one clinic has closed
and its resources used to bolster another clinic to better meet needs of women in
Dublin.
The routine taking of cervical smears from every woman who has been referred to the
clinic is unnecessary and a waste of resources. The colposcopist should have access to
the Register for the results of referral smears so that clinical appearances of the cervix
can be properly interpreted. Repeat smears at the first visit should only be required in
those few women with an abnormal referral smear who show no colposcopic
abnormality.
7.2.7
Histopathology Service
Mid Western Regional Hospital Histopathology Department was the diagnostic service
provider to Phase 1 ICSP. The Department of Health and Children fund the laboratory
service providers through the acute hospital services. A local committee drives
coordination with the ICSP.
33
7
evaluation of phase one of the icsp in the mid-western
health board area
Although the ICSP QA lead for cervical Cytopathology doubles for histopathology, no
standards have been agreed for the Programme histopathology services, and there
have been no ICSP QA visits to date.
While there is a specialist cervical histopathologist in Limerick, in most histopathology
laboratories the reporting of cervical biopsies is shared among all pathologists with no
single pathologist taking a specialist interest in cervical pathology - particularly with
respect to liaising with the Programme and providing quality assurance data. It is
worth noting that histopathology was the last set of standards published by the NHS
Cervical Screening Programme. These were published in 1999 (NHSCSP Publication No
10), three and a half years after the first publication of standards for cytopathology
laboratories. NHSCSP histopathology standards are now in place in the UK.
However, there has been some progress. SNOMED coding has been agreed for all
cervical biopsy results and the interface between Limerick Histopathology Laboratory
and the Register Office has been established successfully. Unfortunately there is a
backlog of results to be coded and sent to the Register which is delaying analyses of
the programme standards. This also compromises failsafe follow up and must be
addressed.
As with cervical cytopathology, the ICSP has encouraged the histopathology
laboratories to seek formal laboratory accreditation with an External Agency such as
CPA(UK)Ltd. As with cytopathology laboratories, only a few histopathology
laboratories have applied for accreditation and none have been inspected to date.
8
a national rollout of the cervical screening programme
35
8.
A NATIONAL ROLLOUT OF THE CERVICAL SCREENING PROGRAMME
There remain many opportunities and challenges for the national Cervical Screening
Programme in Ireland. The UK NHS CSP concentrated on improving coverage and
failsafe systems in its first screening round (1988-1993) and on improving quality in the
next screening round. The challenge for Ireland is to achieve both in the first screening
round. The programme must be high quality but affordable, be shown to give value for
money and give equal access to all eligible women.
8.1
Preparedness for a national rollout and the critical success factors
Experience from the first phase of the Irish Cervical Screening Programme in the
Midwest is that a number of issues need to be resolved, namely:
8.1.1
Governance Structures, Management and Accountability
A home must be found for the ICSP in the current health reorganisation. I recommend
it is based within Population Health since the natural home for a cervical screening
programme is within a Population Health department. Given that the health service in
Ireland is undergoing restructuring at present and the final structure is not yet clear to
me, it is difficult for me to recommend exactly where the Irish Cervical Screening
Programme should be located. However, wherever it is based, it is critical to ensure
that the ICSP is appropriately organised and governed to achieve success.
A management structure with clear governance arrangements is required to ensure
that the legal obligations to eligible women, health care professionals and staff
participating in the Programme including Phase One are met.
The establishment of effective governance arrangements with immediate effect is
necessary in order to provide overall leadership and direction in terms of policy
development, standards, accountability and achievement of value for money for Phase
One and any extension of the programme.
A key element to be addressed in terms of clarity with regard to governance for the
Programme involves the synergies with Breastcheck in moving towards a public health
population based screening programme model within the health reform process.
A robust programme model is required to allow for national expansion, one that would
fulfil the criteria for a national public health programme and lead to a reduction in
morbidity for Irish women. All international evidence shows that population screening
programmes are cost effective when organised as a Public Health Programme,
underpinned by clear standards, managed clinical networks, clear accountability across
organisational boundaries, clearly documented individual and organisational roles and
responsibilities (defined in written protocols) as well as a dynamic policy and quality
assurance system. The model of service delivery should facilitate the required
accountability to ensure a quality assured screening service. A single central
Programme office with clear lines of policy that everyone adheres to, responsibility,
ring fenced funding and multi-annual budget planning is required.
At the current time, the Programme works within existing services. Integration with
service providers has proved difficult because of competing priorities. This has led to a
serious lack of accountability to the Programme. A contract of service for the
programme must be agreed before programme work is undertaken. The hospital could
bid into the programme for this or the Programme could directly contract the staff and
service. Consideration could be given to the Programme establishing independent
structures for service delivery in contracting its own staff and facilities particularly for
the cervical cytopathology laboratory service.
There are no structures in place to allow immediate review or revision of the QA
document by a professional group. Given Ireland’s geographic location and population
8
a national rollout of the cervical screening programme
size and since the Irish cytology and colposcopy services already aspire to the UK
based NHSCSP and BSCCP standards, the ICSP should immediately adopt these
standards as the minimum with review by the ICSP QA Steering Committee in the light
of the experience of the first three years of invitational screening. Phase 1 ICSP has
already aligned in an observer status at the NHSCSP QA Committee.
QA visits should be completed and any resulting recommendations adopted. Formal
registration of smear takers should be implemented.
8.1.2
Professional Direction - National Committees
A multi-disciplinary National ICSP Steering Committee with representation from
management and the professional academic bodies is needed to effectively direct
policy development and issues pertaining to the Programme to reduce risk. A National
ICSP Quality Assurance Committee with external representation is needed to adopt
and review standards. Both committees should have a 2-3 year timeframe.
Consideration should be given to requesting nominees from professional academic
bodies. These individuals should represent their institutions and have delegated
authority to make decisions on behalf of their professional body where appropriate.
A clearly defined screening policy should be developed and implemented which should
be subject to ongoing review in the light of developing evidence, guidance, technology
and government policy. Processes should be put in place to evaluate new technologies
in cervical screening with respect to the ICSP such as computer assisted microscopy
and biomarkers as well as HPV testing.
8.1.3
Programme Direction
The post of Executive Director whose job plan should include the management
responsibilities for adoption of a business-model approach for the Programme should
be established. The Director would, when necessary, enter into contract arrangements
with service providers within existing structures to meet the needs of the Programme
or when appropriate to enter into arrangements with private individuals or agencies.
Accountability from primary care and hospital services (laboratories and colposcopy
clinics) will require review of existing and some new contracts in the light of incoming
health reforms. Ring fenced funding and direct sessional contracting for the Irish
Cervical Screening Programme would assist in accountability in its dealing for service
delivery in the hospital sector.
The Director should report to the Director of Population Health and would be advised
by the Steering Committee for policy and to the Quality Assurance Committee for
standards development.
8.1.4
The Population Screening Register
A complete Register of all women in Ireland aged 25 – 60 years using their unique PPS
identifier must be established. The central cervical screening Register information
system does not currently hold all the relevant clinical history for each woman. A
methodology of recording on the Register all previous smear / colposcopy history for
women joining the programme must be agreed and implemented as soon as possible.
The issue of consent requires to be explored with a view to enabling the Programme to
access existing cytology laboratory databases. Responding to an invitation or direct
referral for a cervical screening test carries with it implied consent for the test and full
buy in to all aspects of the programme.
a national rollout of the cervical screening programme
37
All database owners should clean up their data as best they can, especially merging
duplicate records for women. This task needs to be as automated as possible and may
need outsourcing. Everyone should co-operate to identify women who have been
screened in the past.
The use of the personal public service number (PPS) (1998 Social Welfare Act) should
be made mandatory for all laboratory requests for programme smears and for smears
from women requesting entry to the Register. This will allow keeping all the cervical
screening records for one woman in one file on the Register and facilitates recall at an
appropriate time as well as failsafe follow up. The introduction of a standard cytology
request form which includes the PPS number will facilitate this and all cytology reports
should include a management recommendation to ensure appropriate failsafe follow
up.
All stakeholders (laboratory, Colposcopy Clinic, Primary Care and the Programme
Register Office) should be able to access a women’s history in order to ensure delivery
of appropriate healthcare. This avoids the necessity for colposcopists to have a printed
copy of the referral smear result for each patient. Laboratories should be able to see
the results of previous smears on the woman reported elsewhere and thus give the
appropriate management recommendation. IT linkages between laboratories,
colposcopy clinics, smear takers and the Register must be implemented and tested.
8.1.5
Uptake of screening
The screening interval should be agreed. I recommend an interval of 3 years for women
aged 25 to at least 35 years (extended possibly up to 45 years if laboratory capacity
allows). Thereafter 5 yearly screening is adequate for women who have 2 consecutive
negative smears and no history of abnormality.
Uptake of screening and entry to the Register must be encouraged together with a
cessation of all ad hoc and opportunistic smears. Repeat smears at the first colposcopy
visit following an abnormal smear result should be stopped and all annual smears
banned except in the follow up for abnormal smears or treatment – including repeat
after first ever smear. This is required to free up laboratory capacity to deal with
programme smears and allow equity of access for all women as well as value for
money.
Promotion to health care professional (primary care, labs, hospital clinic staff,
colposcopy, practice nurses, GUM, maternity services, gynaecologists,) including
realistic time scales for implementation and recommendations of what should happen
in the meantime. The transition period timescale before first round invitation letters
are sent out should be agreed and routine activity except for repeat opportunistic
smears should continue in parallel.
Since women can enter the Programme by responding to an ICSP invitation letter or by
direct referral at the discretion of the clinically responsible doctor, consideration should
be given to a rapid national extension with conversion of initial opportunistic smears
to Programme entry policy. Subsequent opportunistic smears however, must be
discouraged.
8.1.6
The screening test
Currently two types of screening test are in use: the conventional Pap smear and
ThinPrep (LBC). A single standard screening test needs to be agreed. I recommend
moving to liquid based cytology since it brings an immediate reduction in the workload
for smear takers, laboratories and colposcopy clinics due to a reduction in
unsatisfactory smears (requiring an immediate repeat) and an increase in laboratory
8
a national rollout of the cervical screening programme
capacity since more slides can be screened each day by laboratory staff. This may
require an EC Procurement process so an early decision is critical. Moving to LBC will
require a national training programme for converting laboratory staff and smear takers
to the new technology and this will take many months to implement – yet another
reason for a rapid decision.
Serious consideration should be given to choosing only one LBC system at present to
allow a more cost effective national procurement. If more than one LBC system is
accepted for implementation it may prove expensive or impossible to organise a
satisfactory national external quality assurance programme for the four laboratories
and for smear takers. Cognisance should be taken of the potential need for sharing
workload between laboratories in the event of one laboratory developing a significant
backlog and to the risk of samples being unsuitable for analysis if collected by a smear
taker using one methodology and sent to a laboratory using the different methodology
since the sample collection methods are not compatible.
Consideration should be given to evaluating the use of computer assisted cytology in
the ICSP since this would further increase laboratory productivity.
8.1.7
Service quality and delivery
There must be a speedy policy decision by the DOHC about the number and location of
cervical cytopathology laboratories, histopathology laboratories and colposcopy clinics
so that there is appropriate capacity for the programme now and in the future. The
cytopathology laboratory, colposcopy clinic and histopathology service should work
together as a team. I recommend services be based in four centres, two in Dublin, one
in the south and one in the west. It may be appropriate for a hub and spoke
arrangement for outreach colposcopy services to be set up so that women do not have
to travel too far for colposcopy investigations and treatment.
All laboratories providing services for the programme should apply for external
laboratory accreditation before January 2006.
The Irish Hospital Services Accreditation Board has been requested by the DoHC (April
2004) to advise on an external accrediting body for the cytology service. A process of
seeking accreditation is essential before the invitational Programme is implemented.
The Health Information and Quality Authority (HIQA) due to be established in autumn
2004, will be addressing standards in all health care providers. I am presuming HIQA
will have the responsibility of declaring External Quality Assurance accreditation for all
screening programme stakeholders and ICSP EQA visits and external accreditation will
fall under HIQA. I would recommend that a laboratory accreditation system similar to
CPA(UK)Ltd is used by HIQA, However, if CPA(UK)Ltd or equivalent laboratory
accreditation is approved, this is a process driven accreditation and there will be
therefore a continuing need for outcome monitoring – for example via ICSP QA visits.
Smears are taken in too many diverse clinical settings and samples are sent to a
diversity of laboratories. This makes call / recall and failsafe follow up difficult. Biopsy
/ Smear correlation (quality assurance) is also impeded.
Training for all personnel providing services to the programme must be given a high
priority and each individual stakeholder must comply with the training and continuing
education standards agreed. Training should be certified by a professional authority
e.g. RCSI / ICGP or equivalent. Training of laboratory staff for LBC should start as soon
as possible and be co-ordinated with smear taker LBC training so that the change from
conventional smears to LBC follows the ability of laboratory staff to evaluate LBC
samples.
a national rollout of the cervical screening programme
39
The fee agreed with GPs for smear taking services Phase 1 ICSP has proved expensive
in terms of the Programme as a whole. Women should have the choice of a female
smear taker and primary care nurses in filling this role may prove to be more cost
effective smear takers. Any review of the GMS contract should include the fee for
Programme smears.
8.2
STEPS IN TRANSITION TO FULL IMPLEMENTATION
There should be a transitional phase lasting no more than 18 months during which all
preparations must be put in place and this- should start immediately. There are big
challenges to be met in this period which is why there must be no delay to starting.
Transition can be followed by first full invitational round of the Irish Cervical Screening
Programme. During the 18 month period all stakeholders should adjust their practices
and prepare themselves for full implementation of the programme. There is a need for
transition to be managed in a professional way with authority to implement changes. I
recommend the appointment of an Implementation Project Manager on a short term
contract to support the Programme Director in the coordination of training to service
providers during the transition phase. The key tasks to be completed in the Transition
period are summarised in Table 11. Significant investment is required to ensure these
tasks are completed in time for national roll out within 18 months.
8
a national rollout of the cervical screening programme
Table 11 Summary of tasks that must be completed during the transition phase.
•
Maintain uptake of screening but cease routine repeat after first-ever smears,
repeat at first colposcopy visit and all other inappropriate annual recall. Prepare
for replacement of ad hoc opportunistic smears with call recall
•
Appoint a Programme Director
•
Submit a business plan and job description for a project manager
•
Appoint a Project manager for the transition period
•
Establish a Steering Committee
•
Agree screening age group and screening interval
•
Implement NHSCSP quality standards as a minimum for all areas of the Programme
•
Agree internal quality control procedures for laboratories
•
Agree move to LBC modality as the single screening test
•
Agree that national Register should hold all relevant clinical information on
women’s screening history
•
Make women’s history on the register available to laboratory staff, GPs and other
Programme personnel
•
Establish confidentiality rules and code of practice for access to register with
regular random audit of access to records
•
Decide the geography of the 4 regional screening centres
•
Decide which laboratories will become the four national screening laboratories
•
Ensure laboratory accreditation, training and capacity planning/ throughput
•
Decide which colposcopy clinics will become the four national centres
•
Ensure colposcopy accreditation, training and capacity planning / throughput
•
Prepare call and recall as focus with mapping of uptake in the four regions
•
Review current funding arrangements and plan for redistribution of resources
where appropriate to ensure affordability and cost effectiveness
•
Establish a training programme for all personnel in the programme
•
Clarify consent issues
•
Consider fees for smear taking within GMS contract.
a national rollout of the cervical screening programme
41
8.3
IDENTIFICATION OF THE IMPROVEMENTS THAT MAY BE NEEDED TO MOVE
FORWARD NATIONALLY.
8.3.1
Cervical cytopathology as service.
A major challenge is the re-organisation of the cytopathology laboratories (internal
and external) to meet the capacity needs of the programme. (See appendix 1 Capacity
calculations) Turnaround times are currently wholly unacceptable in many laboratories.
Women should receive the result of their screening test within four working weeks.
Laboratory throughput must be increased. A defined productivity standard per medical
laboratory scientist would assist capacity planning based on projected workload for
the catchment area.
Cervical cytopathology requires to be rationalised to four national screening cytology
laboratories with specific catchment areas defined by the primary care practices
registered with the Programme laboratory on a county basis. Identification of
laboratories must be on the basis of seeking external accreditation and meeting ICSP
(NHSCSP) QA standards. Laboratories should have a minimum annual throughput of
25,000 ideally moving to 80,000 within a short timescale. Only Programme policy
smears should be processed at the national screening laboratories.
During the18 months transitional phase, it is likely that the national screening
laboratories will require the assistance of other currently existing cytopathology
laboratories to help manage the conventional smear workload while the screening
laboratories are training to convert to liquid based cytology. Work could be allocated
to existing cytology laboratories or to any other cytology laboratory on the basis of
that laboratory’s ability to deliver the specified service to the required level of quality.
Once the invitational programme is established, some of the screening laboratories
may find it necessary to consider a short-term hub and spoke model whereby all
programme samples are sent to a central slide preparation facility for processing. Some
of the slides would be forwarded to a peripheral laboratory for microscopic
assessment. The screening laboratory would be responsible for all data management
and quality assurance. This centre facility would be the nucleus for a single reporting
laboratory in due course.
Retention of staff might benefit from the career progression that would result from the
introduction of the Advanced Practitioner in Cervical Cytopathology grade of Medical
Laboratory Scientist. After a rigorous training programme and passing an examination,
Advanced Practitioners are able to report abnormal smears and give clinical advice
under consultant supervision. They are a very cost effective grade of staff since they
can dedicate all of their time to the cytopathology service and thus the number of
consultant pathologist sessions required can be markedly reduced. A staffing structure
for the national screening laboratories requires to agreed – a model staffing structure
is presented in Table 12
Standard protocols for internal quality control should be agreed and implemented to
avoid unnecessary work within the laboratory. The responsibility for recalling women
at the appropriate interval should be passed to the Register thus freeing up laboratory
time for assessment of slides. Call and recall should be focused with mapping of
uptake in regions and sub-regions.
8
a national rollout of the cervical screening programme
8.3.2
Colposcopy service as service provider for the Programme.
Colposcopy should be an investigation, diagnosis and treatment service based in an
outpatient setting – and women attending should not be considered as patients unless
an abnormality if identified. Colposcopy services should be rationalised to 4 national
centres aligned to cervical cytopathology and histopathology services with
geographically spread satellite clinics for access to the population. Satellite clinics
should be managed centrally by the national colposcopy centres and quality standards
included in those of the national centre.
Each clinic should have a Lead consultant in charge of the service with a dedicated
colposcopy sister and dedicated clerical staff using the colposcopy information system.
The employment of suitably qualified Nurse Colposcopists should be considered. Local
hospital management should be responsible for the service based on being
accountable to ICSP QA standards.
Regular clinicopathological conferences (multidisciplinary meetings) should be held
with review and discussion of discrepancies between cytopathology and colposcopy or
biopsy findings. Case conferencing for difficult cases should be the norm.
Table 12 Model structure of a national cytology screening laboratory
•
At least two consultant pathologists to provide between them a total of 13 sessions
per 25,000 cervical smears per annum
•
A named lead consultant histopathologist with an interest in cervical cytology minimum 7 sessions per week to cervical cytology with responsibility for the quality
of the service and accountable to the Director of the ICSP
•
At least one other consultant histopathologist with an interest in cytology minimum 5 sessions per week to cervical cytology
•
If an Advanced Practitioner grade of staff is employed then total histopathology
consultant minimum time could be reduced to 5 sessions between the two
histopathologists
•
A chief medical laboratory scientist to provide medical laboratory scientist
professional leadership and other defined responsibilities including liaising with the
ICSP
•
Access to a laboratory Quality Manager for accreditation purposes (the QM
sessions can be shared with other laboratory specialties)
•
A Training Officer - a senior medical laboratory scientist to organise training and
QA (should also participate in checking and / or primary screening)
•
Primary screening – 1 basic grade medical laboratory scientist or above per 5,000
conventional smears or 7,000 LBC samples
•
Checking - 1 senior medical laboratory scientist per 2-3 primary screeners
•
Laboratory aid grade staff may be employed for sample preparation
•
Clerical staff will be required not only for data entry and result generation but also
for IT linkage requirements for the ICSP.
a national rollout of the cervical screening programme
43
8.3.3
Histopathology service
There is a need to rationalise services for cervical biopsies taken at colposcopy clinics.
There should be four national histopathology centres aligned to cytology and
colposcopy services. Good practice dictates that the histopathologist reporting a
cervical biopsy should have access to the referral smear result and, where there is a
discrepancy between the biopsy and the cervical smear results, the actual cytology
slide should be available for review. Thus the histopathology laboratory must work
closely with the cervical cytopathology laboratory with which it is aligned and with its
colposcopy clinic staff. A lead histopathologist for cervical biopsies should be
identified with responsibility for liaising with the ICSP. Consent should be sought at the
time of surgery for all histopathological reports for cervical biopsies and hysterectomy
specimens to go on the Register.
All biopsy results should be coded using SNOMED or the ICD classification. If the local
hospital management continues to be responsible for the service it must be
accountable to the ICSP for the QA standards. Hysterectomy samples will continue to
be reported in other laboratories but the cervical SNOMED codes should be made
available to the Register.
There should be a process of correlation with colposcopy and cytology including an
audit of invasive cancers. The histopathology service should adopt the NHSCSP
standards as a minimum immediately.
appendix 1
LABORATORY CAPACITY IMPLICATIONS OF PROPOSAL
TO SCREEN 25-34 YEARS THREE YEARLY
AND 35-60 YEARS FIVE YEARLY
Using 2002 CSO CENSUS FIGURES
FEMALES IN IRELAND (IN THOUSANDS)
25-34
308.9
35-44
283.1
45-54
238.9
55-64
174.2
Grand total
1005.1
25-34 yr olds: 308,900 - 3yearly is 103,000 at 100% and 82,400 at 80% annually
35-64 yr olds: 696,200 - 5yearly is 139,240 at 100% and 111,392 at 80% annually
Total annual capacity required is: 193,792 at 80% annually
Limitations:
• CSO statistics go beyond the 25-60 age group to include those up to 64 years and
are therefore an overestimate
• Cytology capacity must include;
- 10% repeats for unsatisfactory tests (conventional) or 2% (LBC);
- 5% repeats for abnormal tests and
- 2-4% (2% international figure - 4% phase 1 evidence) post colposcopy
surveillance
Therefore, additional annual capacity of 19% (worst-case scenario) to 9% (best-case
scenario) is required.
25-34 yr olds: 3yearly
= 82,400 at 80% of target group annually
Plus 9% (7,416)
= 89,816
Plus 19% (15,656)
= 98,056
35-64 yr olds: 5 yearly = 111,392 at 80% of target group annually
Plus 9% (10,025)
= 121,417
Plus 19% (21,165)
= 132,557
Grand total annual cytology capacity requirements 25-64 year age group in this
instance is from: 211,233 to 230,613 assuming 80% coverage.
45
LABORATORY CAPACITY IMPLICATIONS OF PROPOSAL
TO SCREEN 25-44 YEARS THREE YEARLY
AND 45-60 YEARS FIVE YEARLY
25-44 year olds = 592,000 - 3 yearly is 197,400 annually at 100% uptake
and 157,920 at 80% uptake
45-64 year olds = 413,100 - 5 yearly is 82,620 annually at 100% uptake
and 66,096 at 80% uptake
Total annual capacity required is: 224,016
Assuming an additional annual capacity of 9-19% as before.
25-44 year olds: 3yearly
= 157,920 at 80% of target group annually
Plus 9% (14,213)
= 172,133
Plus 19% (30,005)
= 187,925
45-64 year olds: 5 yearly
= 66,096 at 80% of target group annually
Plus 9% (5,949)
= 72,045
Plus 19% (12,558)
= 78,654
Grand total annual cytology capacity requirements 25-64 year age group in this
instance is from: 244,178 – 266,579
CURRENT AND FUTURE LABORATORY CAPACITY.
In 2002, a total of 230,000 conventional smears were processed in the Irish
laboratories. 70% were performed in 5 laboratories. Moving to LBC and streamlining
internal quality control practices should allow an additional throughput of at least
50% i.e. up to 345,000 capacity across all laboratories.
47